Version July 2025
INTRODUCTION —
Allogeneic hematopoietic stem cell transplantation is the only widely available curative therapy for thalassemia (although autologous transplantation using gene therapy or gene editing are close to becoming available). Transplantation carries many risks and costs, although it is much less expensive than gene therapy. Nevertheless, many individuals will not have a suitable donor, sufficiently resourced care facilities, and/or financial resources to pursue transplant.
This topic discusses transplantation in patients with transfusion-dependent thalassemia (beta thalassemia major), including predictors of a successful outcome, pretransplant evaluation, donor selection, and source of stem cells, as well as management issues in the immediate post-transplant period.
Care following transplant, including donor chimerism (ie, engraftment), graft failure, retransplantation, chronic graft-versus-host disease, and reduction of excess iron stores, is discussed in detail separately. (See "Thalassemia: Management after hematopoietic cell transplantation" and "Iron chelation: Choice of agent, dosing, and adverse effects".)
The general management of thalassemia, including the decision to pursue medical therapy or transplantation, is also presented in detail separately. (See "Management of thalassemia".)
INDICATIONS AND PREDICTORS OF A GOOD OUTCOME —
Allogeneic transplantation carries many risks and costs, and for many individuals, a human leukocyte antigen (HLA)-identical donor is not available. Thus, the decision to pursue transplantation is challenging and must incorporate information about the patient's age, health, and resources as well as the values and preferences of the patient and family/caregivers.
Indications — Transfusion dependence due to thalassemia is the principal indication. The ideal candidate is a child who has been treated with regular transfusions and chelation and who has an HLA-identical donor; such a patient has a more than 90 percent likelihood of cure and a 4 percent risk of transplant-related mortality. For such patients, we suggest allogeneic hematopoietic stem cell transplantation (HSCT) rather than lifelong medical therapy (algorithm 1).
Childhood is the optimal age; the procedure should be performed as early as possible to avoid complications associated with chronic transfusions and iron toxicity [1]. However, there is no absolute age cutoff if the patient is healthy and can tolerate the procedure. Similarly, the lack of a sibling donor is not an absolute contraindication, as fully matched unrelated donor (MUD) transplantation has been associated with similar success rates in experienced reference centers.
The algorithm illustrates the role of HSCT in thalassemia management (algorithm 1).
●Allogeneic HSCT has been studied extensively for beta thalassemia major, with excellent outcomes in selected patients, as discussed below and in the following section. (See 'Outcomes: allogeneic transplantation versus medical therapy' below.)
●Allogeneic HSCT for severe alpha thalassemia has been described in very few case reports either performed in utero or immediately after birth [2]; homozygous alpha thalassemia causes hydrops fetalis and is not compatible with survival in utero or soon after birth.
The likelihood of cure decreases and the risk of death increases with older patient age, longer iron toxicity exposition, and a less well-matched donor. Because the risks of HSCT and medical therapy are very different, the values and preferences of the patient and family/caregivers, availability of the procedure, and costs play a large role in deciding whether or not to pursue transplant. (See "Management of thalassemia", section on 'Decision to pursue allogeneic HSCT'.)
HSCT outcomes are best for individuals who have undergone rigorous medical therapy with transfusions and iron chelation, which minimizes the risks of organ damage from iron toxicity. (See 'Outcomes: allogeneic transplantation versus medical therapy' below.)
While it may be possible to pursue transplantation in individuals with extensive iron stores, a previous episode of heart failure, or compensated cirrhosis, there is a greater likelihood of adverse outcomes in such individuals, and these increased risks may impact decision making. We generally would not pursue this approach for a patient with severe organ damage from iron overload (eg, uncompensated cirrhosis).
For adults who have undergone lifelong regular and consistent chelation therapy, outcomes from HSCT and medical therapy are less well documented [3]. We generally would favor continuation of medical therapy in such an adult. However, adult transplant data are derived from older studies in which participants had previously received suboptimal transfusion and chelation regimens. In the modern era, it is likely that adult thalassemia patients are in much better general clinical conditions and, due to the concept expressed above, transplant results could be better than in previous years and a controlled transplant program could be developed [1].
For children or adults who have not received regular lifelong transfusion and chelation, the risks and benefits are less certain, and decisions are highly individualized. For individuals with comorbidities such as kidney failure, we would avoid transplantation because the risks of transplant toxicity would be too great.
Our approach is consistent with 2014 guidance from an international expert panel from the European Society for Blood and Marrow Transplantation (EBMT), of which one of the authors (EA) is a member [1].
Outcomes: allogeneic transplantation versus medical therapy — HSCT and medical therapy both have risks. Transplantation has a higher up-front risk of transplant-associated morbidity and mortality, but most of the patients who survive are cured of thalassemia. Medical therapy must be continued throughout life, and it carries risks associated with chronic transfusions (alloimmunization, transfusion reactions) and chelation therapy (neurotoxicity, kidney toxicity, and others). Each approach has distinct long-term outcomes that should be taken into account.
There are no randomized trials comparing allogeneic HSCT with medical therapy for transfusion-dependent thalassemia [4].
Outcomes for both modalities continue to improve with greater experience and refinements in management. Thus, assessment of outcomes generally must rely on the experience of transplant centers and comparisons across studies.
●Allogeneic HSCT – Thousands of patients with thalassemia have undergone allogeneic transplantation worldwide. Outcomes were illustrated in a 2016 retrospective series from the EBMT that included data from 1493 patients with beta thalassemia major transplanted after year 2000 [5]. Overall survival for the whole cohort at two years was 88±1 percent. Features that correlated with improved survival included:
•Younger age – Younger patients had better outcomes than older patients, with an age threshold of 14 years or younger determined to be optimal. For the entire cohort, 1359 transplants (91 percent) were done in children, with a median age of 6.6 years. Two-year survival was 96 percent for those transplanted below age 14 years and 82 percent for those transplanted above the age of 14 years.
•Matched sibling donor – HLA-matched siblings were associated with better outcomes than partially matched related donors or MUDs. For the entire cohort, 1061 transplants (71 percent) were done with HLA-matched sibling donors. Two-year survival was 91 percent in the HLA-matched sibling group, 68 percent with a mismatched related donor, and 77 percent with a MUD. Similar results have been reported from China and India [6-8].
•Matched sibling umbilical cord blood or bone marrow – Umbilical cord blood or bone marrow from an HLA-identical sibling was associated with better outcomes than peripheral blood stem cells. Two-year survival was 93 percent with umbilical cord blood, 90 percent with bone marrow, and 81 percent with peripheral blood stem cells, a statistically significant difference.
For all of these groups, approximately 6 to 8 percent of patients who survived the transplant had recurrence of thalassemia (eg, due to graft rejection or graft failure), making thalassemia-free survival approximately 6 to 8 percent lower than overall survival [5]. The effects of recipient age and relatedness of the donor in this EBMT report are summarized in the table (table 1).
These outcomes represent significant improvements since HSCT for thalassemia was first pioneered in the early 1980s, and results continue to improve over time (figure 1) [9-15]. In a longitudinal series from 2023, 73 patients received a transplant; of these, 10 people (14 percent) died of complications of the transplant [16]. Analysis by birth cohort demonstrated that all of the deaths occurred in individuals born earlier; of great relevance, there were no transplant-related deaths after 2010.
Earlier studies established the importance of pretransplant risk assessment for iron overload, and improvements in chelation therapy and transplant techniques (conditioning regimen and supportive care) are likely to account for the improved outcomes. (See 'Impact of iron overload (Pesaro prognostic system)' below.)
Attempts to reduce mortality by lowering the dose of cyclophosphamide in the conditioning regimen have improved survival, but the risk of thalassemia recurrence was also higher. (See 'Myeloablative conditioning regimens for allogeneic transplantation' below.)
Several centers worldwide have developed expertise in transplanting patients with hemoglobinopathies, and we advise transplantation in an experienced center that has the necessary facilities for a hemoglobinopathy patient available. Outcomes may also be better if care is provided by a specialized thalassemia center.
●Medical therapy – Medical therapy (regular transfusions with iron chelation) has transformed thalassemia major from a lethal childhood disease to a chronic condition with survival into adulthood. Studies have documented progressively better survival over time, likely due to increased monitoring of iron stores, earlier and better use of iron chelation, and the availability of better-tolerated iron chelating agents.
•A 2017 study involving 454 patients with thalassemia major treated with medical therapy in Taiwan reported a decrease in the annual mortality rate from 2.9 percent in 2007 to 0.7 percent in 2011; the oral iron chelator deferasirox became available in Taiwan in 2007 [17].
•A 2008 study from the United Kingdom Thalassaemia Register found a decline in the mortality rate from 12.7 deaths per 1000 patient-years from 1980 to 1999 to 4.3 deaths per 1000 patient-years from 2000 to 2003 [18]. This was attributed to reduced deaths from iron overload, which declined from 7.9 to 2.3 deaths per 1000 patient-years. A modeling study that extrapolated from these and other data predicted a survival rate of 63 percent at the age of 50 years in resource-rich countries [19].
•A 2006 study involving 284 patients with transfusion-dependent thalassemia from Cyprus who were born after 1974 and treated medically found estimated survival rates of 100, 99, and 93 percent at 10, 20, and 30 years of age, respectively [20].
•A 2004 study involving 720 patients from Italy who were born after 1960 and were treated with medical therapy reported that 68 percent were alive at the age of 35 years; survival improved progressively with each birth cohort [21].
Several of these studies have reported better survival outcomes in females than in males [21]. This may reflect better adherence to chelation therapy in females.
As noted above and in more detail separately, survival with medical therapy continues to improve and may exceed that described in the above reports, provided patients are followed in experienced centers and resources are widely available [22]. (See "Management of thalassemia", section on 'Prognosis'.)
The excellent data reported in Italy refer to patients followed in a few reference centers; they represent only approximately 14 percent of the entire Italian thalassemia population. Results reported outside reference centers are much less satisfactory [22].
Impact of iron overload (Pesaro prognostic system) — Outcomes correlate with the exposure to iron, including both the severity of iron overload and the duration of exposure [23]. (See 'Outcomes: allogeneic transplantation versus medical therapy' above.)
The importance of pretransplant iron chelation must be emphasized. It is not simply a matter of chelating iron immediately prior to transplant; chelation should be ongoing throughout the entire period of transfusions to prevent iron-induced organ injury. (See "Management of thalassemia", section on 'Assessment of iron stores and initiation of chelation therapy' and "Iron chelation: Choice of agent, dosing, and adverse effects", section on 'Underlying medical conditions'.)
In the 1980s, the Pesaro group developed a simple prognostic system based on indicators of iron overload to predict outcomes from allogeneic HSCT [13]. The system includes three risk factors that were found to correlate with survival on multivariate analysis:
●Quality of chelation therapy before transplantation (regular versus not regular)
●Hepatomegaly (liver edge palpable more than 2 cm below the costal margin)
●Any degree of liver fibrosis at pretransplant hepatic biopsy examination
Class I patients have zero risk factors and are defined by optimal control of iron stores for their entire lives. Class II patients have one or two risk factors, and class III patients have all three risk factors [13,24].
Data from the Pesaro group on transplantation outcomes in children <16 years of age with thalassemia correlated with survival at three years as follows [13-15,25]:
●Class I – 94 percent
●Class II – 80 percent
●Class III – 61 percent
Although much time has passed since these data were acquired, and these precise numerical classifications are probably outdated, these results are still important because they provide the proof of concept of how prolonged exposure to iron toxicity can be the cause of the oxidative damage to human tissues, which are consequently made more susceptible by transplantation and its complications and toxicity [23].
As an example, in a retrospective study that evaluated outcomes in patients with graft-versus-host disease (GVHD) following allogeneic HSCT, GVHD incidence and severity were similar in the three categories of patients, but the probability of surviving GVHD differed significantly. Mortality rates following acute grade III to IV GVHD were 27, 48, and 84 percent for low-, intermediate-, and high-risk patients, respectively [26].
Considerations for adults — The vast majority of experience with allogeneic HSCT for thalassemia is with children.
In the 2016 report from the EBMT group, adults constituted 9 percent (133 of 1493) of thalassemia HSCT recipients [5]. The median age was 22.9 years (range, 18 to 45 years). At two years of follow-up, adults who received a transplant form an HLA-identical sibling donor had an overall survival of 80 percent and a thalassemia-free survival of 76 percent.
As with children, prevention of organ toxicity from iron overload is likely to be a key predictor of good outcomes, although this has not been demonstrated in clinical trials. At the time of this analysis, all adults were not regularly receiving chelation therapy and had already developed significant organ injury before regular chelation was initiated, blunting the benefits of iron removal. Controlled trials in well-chelated patients are warranted [23].
Alpha thalassemia — Severe alpha thalassemia presents in utero, and many patients do not survive until birth due to severe hydrops fetalis. However, cases have been reported describing early transfusional support followed by allogeneic transplant [2]. Transplantation is considered investigational in this setting. (See "Alpha thalassemia major: Prenatal and postnatal management".)
PRETRANSPLANT EVALUATION
Assessment of organ damage — The main purpose of the pretransplant evaluation is assessment of the patient's transplant-associated risk due to organ damage from excess iron stores.
We perform the following testing:
●Liver iron assessment – Liver iron is assessed using liver biopsy or magnetic resonance imaging (MRI). This can be omitted in children age 3 years or younger since these children are unlikely to have developed cirrhosis. Compensated cirrhosis (eg, Child-Pugh class A) is not a contraindication to transplantation (table 2), but its impact on prognosis and transplant outcomes must be factored into the decision to pursue a transplant. Conversely, a study documented the possibility of reversal of early cirrhosis and restoration of normal liver function after successful transplantation and complete removal of iron [27]; this finding may impact decision-making. (See "Cirrhosis in adults: Overview of complications, general management, and prognosis", section on 'Compensated cirrhosis'.)
●Testing for viral hepatitis – Viral hepatitis is not a contraindication, but we routinely test for hepatitis viruses (hepatitis B virus [HBV] and hepatitis C virus [HCV]). Eradication of HCV may be appropriate prior to transplantation. (See "Overview of the management of chronic hepatitis C virus infection".)
●Cardiac assessment – Cardiac function is assessed using electrocardiography and echocardiography. Cardiac T2* MRI could be useful in selected cases. Additional testing such as stress echocardiography or ambulatory electrocardiograph monitoring (eg, Holter monitor) may be used to identify more subtle cardiac involvement [28]. A previous episode of iron-related heart failure is not a contraindication, provided the patient has fully recovered and has received adequate chelation therapy.
●Splenic function – Hypersplenism can complicate the post-transplant course by increasing cytopenias and transfusion requirements. In this situation, the decision to perform splenectomy before transplantation is based on individual patient evaluation and center procedures; no randomized trial data are available. In a small retrospective study involving 29 patients with beta thalassemia who underwent pre-transplant splenectomy, five-year survival was inferior compared with a control group of patients with high risk who did not undergo splenectomy (40 versus 72 percent), suggesting that splenectomy should not be routinely pursued [29]. Most of the difference was due to peritransplant infectious deaths [30]. Individuals who do undergo splenectomy should have appropriate vaccinations, as discussed separately. (See "Elective (diagnostic or therapeutic) splenectomy", section on 'Preoperative considerations' and "Prevention of infection in patients with impaired splenic function", section on 'Vaccinations'.)
●Endocrine function – The endocrine evaluation is performed in patients older than 10 years of age and directed at finding endocrine deficiencies caused by iron deposition in the pancreas, thyroid, and pituitary gland. We obtain a fasting blood glucose level, thyroid function tests, and growth-hormone releasing hormone (GHRH) stimulation test. These evaluations do not impact transplant outcomes or procedures but can be very useful for long-term post-transplant care.
Fertility — Infertility due to temporary or permanent hypogonadism is common following allogeneic transplantation in females and males. We advise sperm banking for post-pubertal males prior to the transplant. For post-pubertal females, it may be possible to preserve ovarian tissue in some cases. (See "Fertility and reproductive hormone preservation: Overview of care prior to gonadotoxic therapy or surgery".)
DONOR SELECTION FOR ALLOGENEIC TRANSPLANTATION —
A human leukocyte antigen (HLA)-identical sibling without thalassemia major is the ideal donor. (See 'HLA-matched sibling donor' below.)
The likelihood of such a donor depends on the number of biological siblings. Approximately one-fourth of the patient's siblings may also have thalassemia major, which eliminates them as possible donors. A donor with thalassemia minor can be an optimal donor, with mild microcytic anemia as the only consequence for the recipient.
For those who do not have an HLA-matched sibling, a matched unrelated donor (MUD) can often be identified from a donor registry. Alternative donor options include the following (table 3):
●In vitro fertilization with embryo selection to produce an HLA-matched sibling without thalassemia who can donate umbilical cord blood. This approach is fraught with ethical, psychological, and legal ramifications and should be pursued, even if legally permitted, only with the appropriate consultations in those disciplines.
●Use of a haploidentical-related donor, such as an HLA-mismatched sibling or parent.
●Use of unrelated donor umbilical cord blood.
●Autologous transplant with gene therapy, which may have increasing importance and wider applicability in the future [31]. (See 'Gene therapy and gene editing' below.)
Use of these alternative donors (those other than a matched sibling or MUD) is considered investigational/experimental by most experts.
HLA-matched sibling donor — An HLA-identical sibling without thalassemia major is the ideal donor because the risk of graft-versus-host disease (GVHD) is minimized. (See "Donor selection for hematopoietic cell transplantation", section on 'Matched sibling donors'.)
An HLA-identical sibling is one who shares the same HLA haplotype at six of six (or eight of eight) HLA loci (HLA-A, B, and DR or HLA-A, B, C, and DR, respectively). The probability that a sibling (assuming they are from the same parents) will be HLA identical is 25 percent, and the probability of having an HLA-identical sibling without thalassemia major is approximately 18.5 percent.
Some parents may have additional children with the hope of having an HLA-matched sibling without thalassemia major. Umbilical cord blood may be collected at the time of birth and tested later. (See 'Sibling cord blood' below.)
It is important to verify that the donor does not also have thalassemia major; this is likely to be obvious in an older sibling but not in a newborn. By contrast, a donor with thalassemia minor can be an optimal donor, with mild microcytic anemia as the only consequence for the recipient.
Matched unrelated donor — A MUD is an unrelated individual who shares the same HLA type with the recipient, typically at 8 of 8 or 10 of 10 loci (HLA-A, B, C, DRB1, and DQB1). Because of the non-malignant nature of this disease, we suggest a 10 of 10 HLA matched unrelated donor. These donors are identified from donor registries. In many cases, ethnic groups with a high incidence of thalassemia tend to be underrepresented in these registries; however, a number of studies have demonstrated good outcomes with MUD transplantation for thalassemia, and experience continues to expand. A typical donor search takes approximately two to four months, and approximately one-third of patients are able to identify a MUD [32].
Although the tested HLA loci from a MUD may appear to be identical to the recipient, a MUD is likely to have antigenic differences that are not detectable by standard testing but are immunologically important. In addition, the majority of MUD searches consider DPB1 when picking a donor. Thus, a matched related donor is preferred to a MUD, and if a MUD is used, molecular HLA typing is used to select the best donor and/or to eliminate donors who would not be appropriate due to molecular HLA mismatches. It has been demonstrated in patients without thalassemia that each allele mismatch reduces survival by approximately 10 percent [33]. (See "Donor selection for hematopoietic cell transplantation", section on 'Unrelated donors'.)
When mismatch occurs, it may be bidirectional or unidirectional. Unidirectional mismatch is typically due to homozygosity of the donor (host-versus-graft direction) or the recipient (graft-versus-host direction) [33]. A study involving 72 patients with thalassemia and their unrelated donors found that the risk of graft rejection and thalassemia recurrence was increased in those with nonpermissive HLA-DPB1 mismatches in the host-versus-graft direction [34].
Even in this setting, relevant clinical improvement occurred, and impressive results have been reported from China in 252 consecutive young children who received a MUD, with 85 percent overall survival and 82 percent thalassemia-free survival [6]. These results have been confirmed in a multicenter Chinese setting [7]. Similar results were reported from India [8].
Embryo selection — For some patients, it may be possible to perform in vitro fertilization (IVF) with preimplantation HLA typing and testing for thalassemia mutations [35,36].
This approach carries a number of ethical, legal, and financial considerations, which are discussed in more detail separately. (See "Preimplantation genetic testing", section on 'Patients who wish to give birth to a child with a compatible HLA type for stem cell therapy of a sibling' and "Collection and storage of umbilical cord blood for hematopoietic cell transplantation", section on 'Ethical and legal issues'.)
Haploidentical donor — A haploidentical donor (also referred to as a mismatched, related donor) is one who shares one haplotype with the recipient, as illustrated in the figure (figure 2).
Experience using haploidentical donors for thalassemia is limited, with results from small series that appear to be improving over time, and the approach is considered experimental in thalassemia by most experts [1].
The following studies illustrate findings from the largest series of haploidentical transplants for thalassemia major:
●A 2020 series described haploidentical transplants in 83 patients with transfusion-dependent beta thalassemia major who received a very complex immunosuppressive preparative regimen starting 70 days before infusion of high-dose T-cell replete peripheral blood stem cells followed by post-transplantation cyclophosphamide on days 3 and 4 [37]. After a median follow-up of 17 months (range 7 to 55), the three-year overall survival and thalassemia-free survival was 96 percent (CI 85.7-98.4).
Engraftment was seen in 29 patients (94 percent), all of whom showed complete donor chimerism. Projected overall survival and thalassemia-free survival at two years were 95 and 94 percent, respectively.
●A 2011 series described haploidentical transplants in 16 patients with beta thalassemia major who received one of two conditioning regimens with infusion of peripheral blood stem cells [38]. Engraftment was seen in 14 patients (88 percent), all of whom showed complete donor chimerism. At a median of 49 months of follow-up, 13 (81 percent) were alive and free of thalassemia.
●A 2010 series described haploidentical transplants using the mother as donor in a series of 22 patients with thalassemia major [39]. Purified stem cells were harvested from bone marrow or peripheral blood with CD3+CD19+ T-cell depletion and add-back. Full engraftment was seen in 14 patients (64 percent). There were two deaths and six cases of graft rejection. The same group presented an update of their data showing a long-term risk of graft failure of 45 percent, resulting in a 10-year thalassemia-free survival of 39 percent [40]. They subsequently modified the T-cell depletion method using T-cell receptor (TCR)-alpha-beta+/CD19+-depleted grafts, which lowered the graft failure rate to 14 percent at four years and improved the thalassemia-free survival rate to 78 percent. We are not aware of any further developments of this approach.
Smaller studies have also produced promising results. As an example, a 2018 series of eight children who underwent allogeneic transplantation using a haploidentical donor described a new conditioning regimen consisting of fludarabine, busulfan, cyclophosphamide, and rabbit antithymocyte globulin (FBCA) [41]. All patients had full donor chimerism. At a median of 36 months, the overall survival was 100 percent, and all were transfusion independent.
STEM CELL SOURCE FOR ALLOGENIC TRANSPLANTATION —
Hematopoietic stem cells (HSCs) and progenitor cells can be isolated from bone marrow, peripheral blood, or umbilical cord blood. As with other benign hematologic conditions for which allogeneic HSC transplant is used, outcomes are generally better using HSCs from bone marrow or umbilical cord blood rather than peripheral blood, likely due to the lower risk of chronic graft-versus-host disease (GVHD).
This is because the HSC product from peripheral blood is enriched in T lymphocytes, which are responsible for GVHD. This effect is tied to a graft-versus-leukemia effect that is beneficial for malignant diseases but not necessary in thalassemia (or other nonmalignant disorders).
However, given the rare but important reports of myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) in gene therapy trials for sickle cell disease, which involve an autologous HSC transplantation, the graft versus leukemia effect widely reported for allogenic transplantation might be important for lowering the risk of bone marrow dyscrasias due to genotoxic ablation of the pre-treatment bone marrow by conditioning agents.
Bone marrow versus peripheral blood — The majority of transplant centers use HSCs derived from bone marrow rather than peripheral blood. The major rationale is to reduce the risk of chronic GVHD; improved overall survival has been noted in some studies but not others, and major differences in graft failure have not been demonstrated.
For patients who undergo allogeneic transplantation for thalassemia, we suggest bone marrow or umbilical cord blood rather than peripheral blood as a stem cell source. This approach is consistent with advice from a 2014 expert panel guideline [1]. Some individuals may use HSCs from peripheral blood in the setting of a clinical trial or if the donor refuses to (or is unable to) donate bone marrow, and it is possible that future changes in conditioning regimens or in GVHD prophylaxis may reduce rates of GVHD.
The lower rate of chronic GVHD with bone marrow stem cells has been demonstrated in several nonrandomized studies involving patients with thalassemia. As examples:
●In a 2007 study involving 183 children who underwent matched sibling donor transplant using bone marrow or peripheral blood stem cells, chronic GVHD was seen in 19 percent of those who received bone marrow and 48 percent of those who received peripheral blood [42]. Acute GVHD was also less common and less severe in the bone marrow group. Two-year disease-free survival and overall survival were similar.
●In a 2010 study involving 52 children with class III disease who underwent matched sibling donor transplant, chronic GVHD occurred in 40 percent of those who received peripheral blood stem cells and 16 percent of those who received bone marrow [43]. There were no differences in overall survival, thalassemia-free survival, or graft failure.
●As described above, a 2016 retrospective series from the European Society for Blood and Marrow Transplantation (EBMT) found a better two-year survival with bone marrow than with peripheral blood as the source of HSCs (90 versus 81 percent), which correlated with lower rates of acute GVHD (3 versus 8 percent) [5].
Umbilical cord blood — Umbilical cord blood is obtained at the time of delivery and may be banked for future use for a related (sibling) recipient or donated to an umbilical cord blood bank for an unrelated recipient. The size of umbilical cord blood units (eg, total nucleated cell count or CD34+ cell dose) must be sufficient to allow for reconstitution of the bone marrow (eg, >3.5 x 107 nucleated cells per kg). However, cell dose has not appeared to have a major impact on outcomes in thalassemia in the setting of a human leukocyte antigen (HLA)-identical sibling [44]. (See "Selection of an umbilical cord blood graft for hematopoietic cell transplantation", section on 'Cell dose'.)
Sibling cord blood — Umbilical cord blood from an HLA-matched sibling is equivalent to or better than bone marrow as a source of stem cells from an HLA-matched sibling. This was demonstrated in the 2016 report from the EBMT, which reported a two-year overall survival of 93 percent in individuals who used umbilical cord blood (all from siblings). (See 'Outcomes: allogeneic transplantation versus medical therapy' above.)
Umbilical cord blood has been associated with lower rates of GVHD than bone marrow in several series [44-46]. As an example, in a 2013 series that included 325 children with thalassemia who received stem cells from a sibling donor using umbilical cord blood or bone marrow, the six-year disease-free survival was similar (80 versus 86 percent), but umbilical cord blood was associated with a lower incidence of acute GVHD (10 versus 21 percent) [44]. However, the children who received umbilical cord blood were younger (six versus eight years), had a better Pesaro risk score (class II to III in 39 versus 44 percent), and were treated more recently.
Unrelated donor cord blood — Unrelated donor umbilical cord blood has not been explored in systematic studies in patients with thalassemia, and we consider this to be investigational therapy. Small studies have described patients undergoing unrelated donor umbilical cord blood transplants. Outcomes have not been very good in some cases (eg, two-year overall survival 62 percent, disease-free survival 21 percent) [47], although they have been better in others (five-year overall survival 88 percent, disease-free survival 74 to 77 percent) [48,49]. Additional clinical trials are awaited.
CONDITIONING REGIMEN —
Conditioning is required to eradicate the recipient's bone marrow and hematopoietic stem cells (HSCs) and to establish a bone marrow microenvironment capable of supporting the donor HSCs (or modified autologous HSCs).
At baseline, individuals with thalassemia have an expanded, hypercellular marrow due to ineffective erythropoiesis. These individuals have an intact immune system and require sufficient immunosuppression to result in sustained engraftment of the allogeneic HSCs.
The myeloablative regimens developed for conditioning have more or less yielded the same overall survival and thalassemia-free survival [50,51].
However, for allogeneic transplantation, different regimens may reduce the risks of graft-versus-host disease (GVHD) or transplant toxicities in certain populations. Addition of fludarabine by several groups in the 2000s has resulted in a reduced risk of graft rejection without additional transplant-related toxicities. Investigational nonmyeloablative and reduced-intensity conditioning regimens are being evaluated, especially in older patients and patients receiving haploidentical transplants.
Myeloablative conditioning regimens for allogeneic transplantation — Standard myeloablative conditioning regimens for allogeneic transplantation typically use chemotherapy alone (without radiation), with the alkylating agents busulfan and cyclophosphamide. Total body irradiation generally is not used because of the effects on growth and development and the risks of radiation-induced secondary malignancies [52]. (See "Preparative regimens for hematopoietic cell transplantation".)
●Busulfan can eradicate the expanded erythropoietic bone marrow but is insufficiently immunosuppressive to allow engraftment of the donor marrow. A typical regimen uses busulfan 16 mg/kg, which can be given orally or intravenously. We typically use intravenous (IV) busulfan at a dose of 3.2 mg per kg per day for four days [1]. Busulfan dosing can eventually be adjusted to a target concentration versus time area under the curve to assure adequate levels with reduced toxicity, but this adjustment is not widely applied [53,54]. Age and liver function do not appear to affect busulfan levels [55].
●Cyclophosphamide was historically added for its immunosuppressive properties but has been replaced by fludarabine, usually at a dose of 40 mg/m2 for four consecutive days.
In high-risk patients, we may modify the conditioning regimen or use a regimen of thiotepa, fludarabine, and IV busulfan for three days. Such a regimen (containing busulfan, cyclophosphamide, thiotepa, and fludarabine) was demonstrated in a 2012 study to reduce the incidence of chronic GVHD [56]. The study included 52 children who received peripheral blood stem cells from unrelated donors and 30 children who received bone marrow from matched sibling donors. The three-year overall survival was 92 percent in the unrelated donor group and 90 percent in the sibling group, and the rates of grade III to IV acute GVHD were 10 and 4 percent; rates of chronic GVHD were not reported. These results have been confirmed in larger series [6,7].
In a 2005 trial that randomly assigned 94 children with thalassemia to receive one of two conditioning regimens (busulfan 600 mg/m2 and cyclophosphamide 200 mg/kg each divided over four days versus busulfan 16 mg/kg and cyclophosphamide 200 mg/kg each divided over four days along with antilymphocyte globulin 30 mg/kg daily for three days), there were no major differences in survival, graft rejection, or transplant toxicities [57]. A lower busulfan trough level correlated with a greater likelihood of thalassemia recurrence.
Other agents have been tested for myeloablative conditioning, such as the busulfan derivative treosulfan (dihydroxybusulfan), the alkylating agent thiotepa, and the antimetabolite (purine analog) fludarabine, which induces profound lymphocyte depletion [50,56,58-62]. Myeloablative conditioning regimens incorporating these agents appear to be highly effective in unrelated donor transplants [56].
Reduced-intensity conditioning regimens (which may still cause myeloablation, with reduced toxicity) are used only in the experimental setting (eg, with haploidentical donors, as described above). (See 'Haploidentical donor' above.)
The use of reduced-intensity conditioning in individuals with greater degrees of iron overload was evaluated in a 2014 study (nonrandomized) that compared results of conventional versus reduced-intensity conditioning in 98 children who were at very high risk (liver size ≥5 cm and age ≥7 years) [63]. Conventional myeloablative regimens consisted of busulfan and cyclophosphamide with or without fludarabine. Reduced-intensity regimens consisted of busulfan plus fludarabine. Overall survival at approximately five years was similar between the standard and reduced-intensity groups (95 versus 90 percent), as was thalassemia-free survival (86 versus 90 percent). However, there were more treatment toxicities in the standard group, including hemorrhagic cystitis and sepsis.
Distinct considerations apply to conditioning regimens for gene therapy.
Nonmyeloablative conditioning regimens — Nonmyeloablative conditioning regimens have been developed for allogeneic transplantation with the hope that engraftment may be possible with less transplant-associated morbidity and mortality. These regimens also appear to be highly effective in inducing stable chimerism with full donor erythropoiesis [1,64]. (See "Thalassemia: Management after hematopoietic cell transplantation", section on 'Mixed chimerism'.)
Despite these advances, nonmyeloablative regimens have not been compared with myeloablative regimens in a randomized trial, and concerns remain about greater risk of thalassemia recurrence with reduced-intensity regimens. Since patients with thalassemia are often heavily transfused and do not receive prior therapy, the barrier to engraftment is higher and graft failure is a major concern with the use of less intensive preparative regimens. (See "Preparative regimens for hematopoietic cell transplantation", section on 'NMA and RIC regimens'.)
GENE THERAPY AND GENE EDITING —
Autologous hematopoietic stem cell (HSC) transplantation using modified autologous HSCs to produce RBCs lacking the thalassemia variant is an alternative to allogeneic transplantation.
The decision to pursue gene therapy for a patient with thalassemia is highly complex. Consultation with experts in thalassemia and transplantation can help in decision-making and understanding evolving criteria, as discussed separately. (See 'Indications' above and "Management of thalassemia", section on 'Decision to pursue allogeneic HSCT'.)
A European Hematology Association (EHA) paper reporting indication and priority criteria for gene therapy in thalassemia has been published [65].
Gene therapy — One approach for beta thalassemia uses viral transduction of a normally functioning beta globin gene into autologous HSCs that are then infused as an autologous HSCT [66]. Myeloablative conditioning is required. Injection of the HSCs directly into bone but with a different cellular product has also been explored, as described in the clinical studies below [67].
●Betibeglogene autotemcel (Zynteglo, Beti-cel)
•Approvals – The gene therapy construct betibeglogene autotemcel (Zynteglo), a lentiviral product containing approximately 24 to 400 million autologous CD34+ cells transduced with a beta globin variant (T87Q), was approved by the European Medicines Agency (EMA) in 2019 for individuals 12 years and older who have transfusion-dependent beta thalassemia with a non-beta0/beta0 genotype (ie, they must have at least one beta+ variant). Because of disagreements about cost coverage, it is not available in Europe despite regulatory approval. It was approved by the US Food and Drug Administration (FDA) for children and adults with transfusion-dependent beta thalassemia in 2022 [68].
•Evidence for efficacy – Several small studies have reported impressive clinical benefits, especially in individuals with some Hb A production at baseline (non-beta0/beta0 genotypes) [31,69]. In the most severe patients, the results were positive but not curative, indicating the high quantitative bar that must be cleared for cure.
-In a study from 2018, 12 of 13 individuals with non-beta0/beta0 genotypes became transfusion-independent, with hemoglobins between 9.2 and 13.7 g/dL [31]. This included nine individuals with compound heterozygosity for hemoglobin E and a beta0 variant. In the nine individuals with a beta0/beta0 genotype or two copies of the IVS1-110 mutation, the median annualized number of transfusions was decreased by 74 percent, and three (33 percent) became transfusion-independent.
-In a study from 2021, 20 of 22 evaluable patients (91 percent) with non-beta0/beta0 genotypes treated with this therapy were transfusion-independent at a median of 29.5 months follow up, including 6 of 7 children <12 years of age [69]. The mean hemoglobin was 11.7 g/dL (range, 9.5 to 12.8 g/dL), mostly consisting of Hb AT87Q (median, 8.7 g/dL; range, 5.2 to 10.6 g/dL).
-A study from 2024 involving 18 patients with various genotypes including beta0/beta0 treated with this therapy reported transfusion-independence in 16 (89 percent) at a median of 47.9 months follow up [70].
•Adverse effects – Toxicities are as expected for autologous transplantation with myeloablative conditioning regimens [31,69,70].
•Concern regarding myeloid malignancy – In early 2021, trials with lentiviral vectors and use of the EMA-approved product were suspended after three individuals with sickle cell disease who were participating in gene therapy trials developed myeloid malignancies (one of these was subsequently determined pre-existing to gene therapy) [71]. The National Heart, Lung, and Blood Institute (NHLBI) paused a clinical study of a different but related lentiviral vector gene therapy program in sickle cell disease "out of an abundance of caution" [72]. Subsequently, investigation of these cases determined that the gene therapy construct was very unlikely to cause AML/MDS, and the trials were resumed. An individual receiving a lentiviral vector gene therapy for adrenoleukodystrophy also developed MDS. Details are presented separately. (See "Management and prognosis of X-linked adrenoleukodystrophy", section on 'Autologous HSCT with ex vivo gene therapy' and "Curative therapies in sickle cell disease including hematopoietic stem cell transplantation and gene therapy", section on 'Concern about myeloid malignancy in gene therapy studies'.)
Myeloid malignancies have not been reported in individuals with thalassemia who are receiving these therapies, but further study is needed to determine the mechanisms of carcinogenesis and how to address them. Contributing mechanisms may include the underlying disease, other medications, changes related to stem cell collection, the conditioning regimen, the viral vector, or others. One case of MDS was reported not to be due to the viral vector [73]; in one case the lentiviral vector was present inside leukemic cells.
•Vector composition – Betibeglogene autotemcel (Zynteglo, Beti-cel) consists of autologous hematopoietic stem and progenitor cells transduced with the BB305 vector. BB305 is a lentiviral vector containing the T87Q variant of the beta globin gene. T87Q is an antisickling variant similar to gamma globin; the distinct sequence that differs from wild-type beta globin allows quantification of the expression level of the variant hemoglobin (Hb AT87Q) from the transgene. This test is not commercially available.
●GLOBE vector with mini beta globin – A study from 2019 described the use of gene therapy in nine individuals with transfusion-dependent beta thalassemia (six children and three adults) [67]. HSCT was performed using myeloablative chemotherapy with direct intra-bone injection of HSCs transduced with the GLOBE lentiviral vector, which encodes a mini beta globin gene with a modified enhancer region [74].
The rationale for intra-bone injection was based on the experience with umbilical cord blood transplantation in malignancies, to overcome the lower number of HSCs in cord blood [75]. Direct injection into bone favors homing of HSCs to bone marrow spaces and avoids the trapping of HSCs in filter organs. With a median follow-up of 18 months, all participants had reduced transfusion requirements, but only three of four evaluated children were transfusion-free at 14, 15, and 19 months. Therapy was well-tolerated with expected mild chemotherapy-related toxicities and no vector-related adverse events.
These observations highlight the promise, limitations, and quantitative challenges of gene therapy. Longer follow-up to monitor the durability of transfected gene activity will be extremely important. Safety concerns are paramount. The BB305 vector has been modified to be replication incompetent and self-inactivating [69]. The modified autologous stem cells must be nonimmunogenic (immunogenicity can result from "leaky" expression of vector proteins) and must sustain long-term expression of the replacement gene. Research that addresses these concerns is ongoing [66,74,76-91]. No further studies are available for this construct, and no application for FDA or EMA registration has been done.
●Alpha thalassemia – No effective means of gene therapy are in advanced stages of development for alpha thalassemia.
Gene editing — Gene editing involves use of molecular techniques to make permanent genetic changes in a cell's endogenous DNA. In contrast to lentiviral gene therapy, which integrates at multiple random locations in each genome and has a potential risk of insertional mutagenesis, gene editing is a non-viral system and does not use a lentiviral vector. Gene editing uses a transient precise editing system with a specific guide RNA.
HSCs can be modified ex vivo and returned to an individual following autologous HSCT with myeloablative conditioning. In theory, it might be possible to inject the HSC gene editing vector directly, if it could be targeted exclusively to HSCs, but this has not been tested clinically. (See "Overview of gene therapy, gene editing, and gene silencing", section on 'Gene editing'.)
Gene editing trials have not been put on hold after the reported cases of myeloid malignancies after gene therapy in sickle cell disease mentioned above. (See 'Gene therapy' above.)
●Exagamglogene autotemcel (Casgevy, Exa-cel) – Exagamglogene uses CRISPR/Cas9 editing to disrupt the BCL11A gene in autologous HSCs, which increases the percentage of fetal hemoglobin (Hb F). This construct was approved for transfusion-dependent beta thalassemia in January 2024 [92].
BCL11A encodes a transcriptional repressor that promotes globin gene switching from gamma globin to beta globin in early infancy, resulting in a decrease in Hb F and an increase in adult hemoglobin (Hb A). Disruption of BCL11A as a means of reactivating Hb F expression is an attractive approach to therapy [93-95]. This is based on the proven clinical benefit by increased levels of Hb F in patients with thalassemia and sickle cell disease patients.
In a 2024 study involving 52 patients with transfusion-dependent beta thalassemia (ages 12 to 35) treated with myeloablative conditioning and autologous transplantation of HSCs treated with this construct reported that among the 35 patients with ≥12 months of follow-up, 32 (91 percent) were transfusion-independent, with a mean total hemoglobin of 13.1 g/dL and a mean Hb F of 11.9 g/dL [96]. There were no deaths and no cancers; toxicities were as expected with myeloablative conditioning.
Evidence for efficacy and safety of this therapy in sickle cell disease is presented separately. (See "Curative therapies in sickle cell disease including hematopoietic stem cell transplantation and gene therapy", section on 'Gamma globin upregulation (including exa-cel, Casgevy)'.)
●Base editing to revert a splice site point mutation – Base editing is a type of gene editing in which DNA is nicked and adenine deaminase is used to convert adenine to guanine (A to G) [97]. This approach could be used to revert a G to A point mutation back to the wild-type G. In a preclinical study, HSCs from patients with the intronic point mutation IVS1-110 (G>A) were base-edited, allowing restoration of beta globin production [98].
●Other gene editing approaches – Use of gene editing to disrupt aberrant splice sites and restore normal beta globin expression is also under investigation [99].
IMMEDIATE POST-TRANSPLANT MANAGEMENT —
Immediate post-transplant management is directed at hematopoietic support, infection prophylaxis, and prevention of graft-versus-host disease (GVHD) in recipients of allogenic transplantations, as discussed in the following sections.
Monitoring for engraftment and chimerism and our approaches to management of graft failure and other transplant complications, including veno-occlusive disease of the liver and transplant-related infections, are presented separately. (See "Thalassemia: Management after hematopoietic cell transplantation".)
Hematopoietic support — Transfusions are typically needed in the immediate post-transplant period. We use cytomegalovirus (CMV)-negative products for CMV-negative recipients, irradiation of all products, and a platelet count threshold of 10,000/microL for platelet transfusion. We do not routinely use hematopoietic growth factors such as granulocyte-colony stimulating factor (G-CSF), although these may be appropriate in selected cases such as in individuals with delayed engraftment. (See "Hematopoietic support after hematopoietic cell transplantation" and "Platelet transfusion: Indications, ordering, and associated risks".)
Graft-versus-host disease prophylaxis — GVHD prophylaxis is used for allogeneic transplant recipients. It is not required for gene therapy or gene editing approaches because the cells are autologous.
We use cyclosporine for GVHD prophylaxis along with a "short methotrexate" regimen consisting of methotrexate (10 mg/m2 on days 1, 3, and 6 after transplantation). An exception is that methotrexate must be avoided after umbilical cord blood transplant due to adverse effects on outcomes [44]. Cyclosporine is continued for one year after transplantation [100]. Because there is no need for a graft-versus-leukemia effect, cyclosporine is tapered slowly during the post-transplant period. For haploidentical transplants (still an experimental option), post-transplant high-dose cyclophosphamide appears to be the most effective regimen.
In transplantation using an unrelated donor or peripheral blood as the source of stem cells, we also add antithymocyte globulin (ATG) to the GVHD prophylaxis regimen. (See "Prevention of graft-versus-host disease", section on 'Introduction'.)
Infectious disease prophylaxis — We use standard antibiotic and antiviral prophylaxis, as discussed in more detail separately. Patients with thalassemia do not appear to have an increased risk of infectious diseases compared with other populations. (See "Prevention of infections in hematopoietic cell transplant recipients" and "Prevention of viral infections in hematopoietic cell transplant recipients" and "Prophylaxis of invasive fungal infections in adult hematopoietic cell transplant recipients".)
Cardiac tamponade — We have observed a higher-than-expected frequency of cardiac tamponade during allogenic transplantations and in the very early post-transplant period. In our 1992 series of 400 patients, eight (2 percent) experienced cardiac tamponade during conditioning or within a month after transplant, and six died of the complication [101]. Pericardiocentesis was the only effective therapy. Similar events have not been reported subsequently. (See "Cardiac tamponade".)
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sickle cell disease and thalassemias".)
INFORMATION FOR PATIENTS —
UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topics (see "Patient education: Beta thalassemia (The Basics)" and "Patient education: Allogeneic bone marrow transplant (The Basics)")
SUMMARY AND RECOMMENDATIONS
●Indications – Allogeneic hematopoietic stem cell transplantation is the only curative therapy for thalassemia. Outcomes are best for individuals treated with rigorous medical therapy with transfusions and iron chelation (figure 1). Other predictors of good outcomes include younger age, matched sibling donor, and hematopoietic stem cells (HSCs) from bone marrow rather than peripheral blood. (See 'Indications and predictors of a good outcome' above and "Management of thalassemia", section on 'Decision to pursue allogeneic HSCT'.)
●Ideal age – The ideal candidate for allogeneic transplantation is a child under 14 years of age who has been treated with regular transfusions and chelation and who has a human leukocyte antigen (HLA)-identical sibling donor. Such patients have >90 percent likelihood of cure and 4 percent risk of transplant-related mortality.
For such patients, we suggest allogeneic transplantation rather than lifelong medical therapy (Grade 2C). Transplant may also be appropriate in older children or adults with an HLA-matched unrelated donor (MUD) who have received appropriate chelation therapy. Medical therapy is a reasonable alternative, using transfusions, and, for older adolescents and adults, luspatercept (algorithm 1). Transplantation and medical therapy differ substantially in their risks and burdens and have not been directly compared. Patient and family/caregiver preferences play a substantial role. (See "Management of thalassemia", section on 'Decision to pursue allogeneic HSCT'.)
●Pretransplant evaluation – The main purpose is assessment of transplant-associated risk due to organ damage from excess toxic iron from exposure. Sperm banking may be appropriate for postpubertal boys. In some cases, oocyte preservation may be possible for girls. (See 'Pretransplant evaluation' above.)
●Donor – A healthy HLA-identical sibling is the ideal donor. A MUD can often be identified from a donor registry. Other options including umbilical cord blood from an unrelated donor and haploidentical transplant are considered investigational but may be appropriate in some cases (table 3). (See 'Donor selection for allogeneic transplantation' above.)
●Stem cell source – For allogeneic transplantation, we suggest bone marrow rather than peripheral blood as a stem cell source (Grade 2C). The major rationale reduced risk of chronic graft-versus-host disease (GVHD). Some individuals may use HSCs from peripheral blood. Umbilical cord blood from an HLA-matched sibling appears to have a lower risk of GVHD than peripheral blood, but it must be banked at delivery. (See 'Stem cell source for allogenic transplantation' above.)
●Conditioning – Standard myeloablative conditioning uses busulfan and fludarabine. Radiation is avoided due to effects on growth and secondary malignancies. Reduced-intensity and nonmyeloablative regimens are under investigation and may be used in high-risk individuals or for haploidentical transplant. (See 'Conditioning regimen' above.)
●Gene therapy – Gene therapy and gene editing approaches are also under investigation; these have enrolled patients in optimal condition. (See 'Gene therapy and gene editing' above.)
●Post-transplantation care – Immediate post-transplantation management involves hematopoietic support, infection prophylaxis, and GVHD prophylaxis. Subsequent care including monitoring engraftment and removing excess iron is discussed separately. (See 'Immediate post-transplant management' above and "Thalassemia: Management after hematopoietic cell transplantation" and "Iron chelation: Choice of agent, dosing, and adverse effects".)
